The present protocol describes the setting up of 3D ‘on top’ cultures of a non-transformed breast epithelial cell line, MCF10A, that has been modified to study Platelet Activating Factor (PAF) induced transformation. Immuno-fluorescence has been used to assess the transformation and is discussed in detail.
Several models have been developed to study cancer, such as rodent models and established cell lines. Valuable insights into carcinogenesis have been provided by studies using these models. Cell lines have provided an understanding of the deregulation of molecular signaling associated with breast tumorigenesis, while rodent models are widely used to study cellular and molecular characteristics of breast cancer in vivo. The establishment of 3D cultures of breast epithelial and cancerous cells aids in bridging the gap between in vivo and in vitro models by mimicking the in vivo conditions in vitro. This model can be used to understand the deregulation of complex molecular signaling events and the cellular characteristics during breast carcinogenesis. Here, a 3D culture system is modified to study a phospholipid mediator-induced (Platelet Activating Factor, PAF) transformation. Immunomodulators and other secreted molecules play a major role in tumor initiation and progression in the breast. In the present study, 3D acinar cultures of breast epithelial cells are exposed to PAF exhibited transformation characteristics such as loss of polarity and altered cellular characteristics. This 3D culture system will assist in shedding light on genetic and/or epigenetic perturbations induced by various small molecule entities in the tumor microenvironment. Additionally, this system will also provide a platform for the identification of novel as well as known genes that may be involved in the process of transformation.
A myriad of models are available to study the progression of cancer, each of them being unique and representing a subtype of this complex disease. Each model provides unique and valuable insights into cancer biology and has improved the means to mimic the actual disease condition. Established cell lines grown as a monolayer have provided valuable insights into vital processes in vitro, such as proliferation, invasiveness, migration, and apoptosis1. Though two-dimensional (2D) cell culture has been the traditional tool to investigate the response of mammalian cells to several environmental perturbations, extrapolation of these findings to predict tissue-level responses does not seem sufficiently convincing. The major limitation of the 2D cultures is that the microenvironment created differs largely from that of the breast tissue itself2. 2D culture lacks the interaction of the cells with the extracellular matrix, which is vital for the growth of any tissue. Also, tensile forces experienced by the cell in monolayer cultures hinder the polarity of these cells, thus altering cell signaling and behavior3,4,5. Three-dimensional (3D) culture systems have opened up a new avenue in the field of cancer research with their ability to mimic the in vivo conditions in vitro. Many crucial microenvironmental cues that are lost in 2D cell culture could be re-established using 3D cultures of laminin-rich extracellular matrix (lrECM)6.
Various studies have identified the importance of the tumor microenvironment in carcinogenesis7,8. Inflammation-associated factors are a major part of the microenvironment. Platelet Activating Factor (PAF) is a phospholipid mediator secreted by various immune cells that mediates multiple immune responses9,10. High levels of PAF are secreted by different breast cancer cell lines and are associated with enhanced proliferation11. Studies from our lab have shown that the prolonged presence of PAF in acinar cultures leads to the transformation of breast epithelial cells12. PAF activates the PAF receptor (PAFR), activating the PI3K/Akt signaling axis13. PAFR is also reported to be associated with EMT, invasion, and metastasis14.
The present protocol demonstrates a model system to study PAF-induced transformation, using 3D cultures of breast epithelial cells, as has been previously described by Chakravarty et al.12. The breast epithelial cells grown on the extracellular matrix (3D cultures) tend to form polarized growth-arrested spheroids. These are called acini and closely resemble the acini of breast tissue, the smallest functional unit of the mammary gland, in vivo15. These spheroids (Figure 1A,B) consist of a monolayer of closely packed polarized epithelial cells surrounding a hollow lumen and attached to the basement membrane (Figure 1C). This process of morphogenesis has been well described in literature16. When seeded on lrECM, the cells undergo division and differentiation to form a cluster of cells, which then polarize from Day 4 onwards. By Day 8, the acini consist of a group of polarized cells that are in direct contact with the extracellular matrix and a cluster of unpolarized cells enclosed within the outer polarized cells, with no contact to the matrix. These unpolarized cells are known to undergo apoptosis by Day 12 of culture, forming a hollow lumen. By Day 16, growth-arrested structures are formed16.
Figure 1: Nuclei of cells in acini stained with a nuclear stain. (A) 3D construction of the acini. (B) Phase Contrast image of MCF10A acini grown on Matrigel for 20 days. (C) The centermost section shows the presence of a hollow lumen. Scale bar = 20 µm. Please click here to view a larger version of this figure.
Unlike 2D cultures, acinar cultures aid in distinguishing normal and transformed cells through apparent morphology changes. Non-transformed breast epithelial cells form acini with a hollow lumen, mimicking the normal human breast acini. These spheroids, upon transformation, show a disrupted morphology characterized by a major loss of polarity (one of the hallmarks of cancer), absence of a lumen, or disruption of the hollow lumen (due to evasion of apoptosis) that may be induced due to deregulation of various genes17,18,19,20. These transformations can be studied using commonly used techniques such as immunofluorescence. Thus, the 3D cell culture model can function as a simple method to investigate the process of breast acinar morphogenesis and breast carcinogenesis. Establishing a 3D culture system to understand the effect of a phospholipid mediator, PAF, will assist in high throughput preclinical drug screening.
This work has adapted the 3D 'on top' culture protocol16,21 for studying transformation induced by PAF22. The phenotypic changes induced by exposure of the acini to the phospholipid mediator were studied using immunofluorescence. Various polarity and epithelial to mesenchymal transition (EMT) markers12,16 were used in the study. Table 1 mentions their normal localization and their expected phenotype upon transformation.
Antibodies | Marks | Normal localization | Transformed phenotype |
α6-Integrin | Basolateral | Basal with weak lateral stain | Strong lateral / Apical stain |
β-Catenin | Cell-cell junction | Basolateral | Abnormal / nuclear or cytoplasmic localization |
Vimentin | EMT | Absent / weak presence | Up-regulation |
Table 1: Markers used in the study. Different markers used with their localization in the presence and absence of PAF treatment.
This method can be best utilized to study/screen plausible drugs and target genes for various breast cancer subtypes. This can provide a drug response data closer to the in vivo scenario, aiding in faster and more reliable drug development. Also, this system can be used to study the molecular signaling associated with drug response and drug resistance.
1. Seeding MCF10A cells in lrECM
2. PAF treatment
3. Re-feeding with fresh media
4. Immunofluorescence study to detect phenotypic changes induced by prolonged PAF exposure
MCF10A cells, upon exposure to PAF treatment, form acinar structures with very distinct phenotypes. α6-integrin was found to be mislocalized with more apical staining. A few acini also showed discontinuous staining (Figure 2A). Both these phenotypes indicate the loss of basal polarity, as evidenced from literature24,25. Earlier reports indicate the controversial role of α6-integrin in cancer metastasis. α6-integrin is present as a dimer with either β1- or β4-integrin. The α6β4 subunit has been found to play a significant role in forming hemidesmosomes in epithelial cells26. The downregulation of this integrin has been found in the prostate and some instances of breast cancers, wherein loss of α6β4-integrin was found in ductal carcinoma of the breast (grade III). The loss phenotype is seen in the cells that metastasize to the parenchyma and pleural cavity27,28,29. GM130 is a cis-Golgi localized protein; Golgi bodies are apically localized in MCF10A acinar cultures16,30. PAF treatment led to the mislocalization of GM130, suggesting apical polarity disruption (Figure 2B). Vimentin is an intermediate filament that is involved in cell migration; it is upregulated when epithelial cells undergo epithelial to mesenchymal transition (EMT)31. PAF treatment of MCF10A acinar cultures led to elevated vimentin levels as observed by the staining intensity (Figure 2C), suggesting EMT.
Figure 2: PAF disrupts polarity and induces EMT-like changes in MCF10A breast acini. MCF10A cells were grown as 3D 'on top' cultures in lrECM. The cultures were treated with PAF on Day 0, 4, 8, 12, and 16. The cultures were maintained for 20 days and then immunostained for α6-integrin (green) and nuclear stain (blue) (A), (B) GM130 (green), a marker for apical polarity, and (C) Vimentin (red), an EMT marker. The centermost stack of the respective acini has been shown. The representative data is for 40-50 acini from three biologically independent experiments. Scale bar = 20 µm. Please click here to view a larger version of this figure.
Established cell line-based models are widely used to study the process of carcinogenesis. Monolayer cultures of cells continue to provide insights into the various molecular signaling pathways that mediate characteristic changes in cancer cells32. Studies on the role of well-known oncogenes such as Ras, Myc, and mutated p53 were first reported using monolayer cultures as the model system33,34,35,36. However, the 2D culture model lacks a lot of important structural and functional parameters present in vivo. Drug screening studies in 2D cultures often show a significant response (cell death or target inhibition), even at lower concentrations, while failing to show any significant effect when admistered in mice37. A major reason for this is the varying drug availability and uptake in a 2D monolayer culture vs. 3D culture38.
The concept of 3D cultures has emerged as a powerful tool for studying the progression of cancer in the last decade. 3D cultures involve growing cells on an extracellular matrix, resulting in the formation of structures that resemble functional units present in the tissue in vivo16. 3D cultures mimic the in vivo microenvironment to a large extent and thus overcome the limitations of 2D cultures39. Moreover, since the cells grown in 3D are of human origin and the system is more amenable to studying early events, it is simpler and more elegant than the rodent models.
Using this platform, a model to study the transformation process following exposure to a phospholipid mediator such as PAF is demonstrated. Studies from our lab have identified that PAF treatment can transform the non-tumorigenic breast epithelial cell line, MCF10A12. This method is optimized to continuously expose the cells to PAF, similar to the in vivo scenario, where PAF is present in the microenvironment.
MCF10A cells, when grown in 2D, have to be passaged every 4 days. If they are not maintained as per the protocol, they behave very differently, which is visible only when grown on ECM16. The passage number of the cells must be kept as low as possible. This method also has limitations due to the variation in protein concentration in different batches of lrECM. This can be overcome by testing the different batches by simply plating cells on a lrECM bed and observing the size distribution of the acini. If there is no large variability in size, the lrECM batch can be considered appropriate for experiments.
Assessment of the transformation is generally done by immunofluorescence, which is demonstrated in the video. While performing immunofluorescence, the steps involving incubation at 4 °C are critical; pipetting out the solution when the lrECM layer is in a liquid state makes the layer uneven, leading to a loss of acinar structures. One of the ways to tackle the unevenness problem is to store the slides at 4 °C on an even platform overnight or change the magnification at which imaging is done. To avoid pipetting out the acinar structures, carefully remove the chamber from the 4 °C storage and remove three-quarters of the volume added initially, then take out the rest during the subsequent washes.
Another major problem faced is the high background due to lrECM. To overcome this problem one can use a 2% secondary blocking antibody instead of 1%. However, increasing secondary blocking does not serve the purpose of staining apical markers and E-cadherin. In such cases, it is advisable to subject the chamber cover glass after Day 20 to PBS-EDTA for 15 min at 4 °C, and then fix the acini following the same protocol as shown in the video. PBS-EDTA partially dissolves the lrECM, thus reducing the background. However, if the incubation time is exceeded, this treatment can result in a loss of structures while pipetting out the solution. Hence, utmost care has to be taken while carrying out this procedure. It has also been observed that storing the slides for longer periods can lead to an increase in background signal. Therefore, imaging must be done as soon as possible. It is advisable to image a minimum of 15 acini per experiment to determine a change in phenotype.
3D acinar cultures have certain limitations, such as difficulty in tracking single cells during live-cell imaging. Certain studies, such as effects on cell cycle, need to be carried out using live-cell imaging to maintain spatial and temporal information. However, due to constant changes in structure, it is difficult to track such parameters in 3D acinar cultures using regular confocal microscopes. This warrants using advanced microscopic techniques such as super-resolution microscopy or Airyscan microscopy. It would also require dislodging of the acinar structures for certain transformation assays such as wound healing, anchorage-independent growth, and in vivo tumorigenicity assays. This dislodging would result in the loss of important spatial and temporal information. Protein lysates collected from the cultures are often diluted due to the presence of IrECM, posing issues with the loading of sufficient amount of lysates. However, this can be overcome to a greater extent by dislodging the acini structures and collecting them before lysis.
Here, a model system has been demonstrated to study phospholipid mediator/immune-associated factor-induced transformation. This model can elucidate genetic and/or epigenetic pathway(s) that may get de-regulated, leading to drastic changes in morphology. This method can also be used to screen potential therapeutics. Such drug screening studies will provide better success than screenings carried out in 2D cultures38. A major highlight of this method is the adaptiveness as per the requirement of the study. The treatment regimen and analysis methods can be modified as needed. Moreover, this method can provide insights into the molecular signaling affected upon treatment40. The technique will also help to predict the possible occurrence of drug resistance. Furthermore, it will also help in elucidating the mechanism involved in drug resistance and identify plausible targets for overcoming the same. Studies have established possibilities of co-culturing multiple cell types in 3D, which can further be modified to accommodate the treatment regimen41. Such opportunities will provide greater insights into the characteristics and molecular changes that happen in vivo during the initiation and progression of breast cancer.
The authors have nothing to disclose.
We thank the IISER Pune Microscopy Facility for access to equipment and infrastructure and support for the experiments. This study was supported by a grant from the Department of Biotechnology (DBT), Govt. of India (BT/PR8699/MED/30/1018/2013), Science and Engineering Research Board (SERB), Govt. of India (EMR/2016/001974) and partly by IISER, Pune Core funding. A. K. was funded by CSIR-SRF fellowship, L.A. was funded through DST-INSPIRE fellowship, V.C was funded by DBT (BT/PR8699/MED/30/1018/2013).
0.05% Trypsin EDTA | Invitrogen | 25300062 | |
16% paraformaldehyde | Alfa Aesar | AA433689M | |
Anti Mouse Alexa Flour 488 | Invitrogen | A11029 | |
Anti Rabbit Alexa Flour 488 | Invitrogen | A-11008 | |
BSA | Sigma | A7030 | |
Chamber Coverglass | Nunc | 155409 | |
Cholera Toxin | Sigma | C8052-1MG | 1 mg/mL in dH2O |
Confocal Microscope | Leica | Leica SP8 | |
DMEM | Gibco | 11965126 | |
EDTA | Sigma | E6758 | |
EGF | Sigma | E9644-0.2MG | 100 mg/mL in dH2O |
F(ab’)2 fragment of antibody raised in goat against mouse antigen | Jackson Immunoresearch | 115-006-006 | |
GM130 antibody | Abcam | ab52649 | |
Goat Serum | Abcam | ab7481 | |
Hoechst | Invitrogen | 33258 | |
Horse Serum | Gibco | 16050122 | |
Hydrocortisone | Sigma | H0888 | 1 mg/mL in ethanol |
Image Processing Software | ImageJ | ||
Insulin | Sigma | I1882 | 10 mg/mL stock dH2O |
lrECM (Matrigel) | Corning | 356231 | |
Mounting reagent (Slow fade Gold Anti-fade) | Invitrogen | S36937 | |
Nuclear Stain (Hoechst) | Invitrogen | 33258 | |
PAF | Cayman Chemicals | 91575-58-5 | Methylcarbamyl PAF C-16, procured as a 10 mg/mL in ethanol |
Penicillin-Streptomycin | Lonza | 17-602E | |
Sodium Azide | Sigma | S2002 | |
Tris Base | Sigma | B9754 | |
Triton X-100 | Sigma | T8787 | |
Tween 20 | Sigma | P9416 | |
Vimentin antibody | Abcam | ab92547 | |
α6-integrin antibody | Millipore | MAB1378 |